Recent Immigrants

Multi-category dot-density maps often work well when the categories cluster geographically. Recent immigrants by country of origin work well for this. ## Data First we grab the immigrant data via cancensus, making use of the CensusMapper API tool to select the regions and variables we need.

#devtools::install_github("mountainmath/cancensus")
library(cancensus)
library(dotdensity)
# options(cancensus.api_key='your_api_key')
regions=list(CMA="59933")
vectors_2011=c("v_CA11N_265","v_CA11N_268","v_CA11N_304","v_CA11N_334","v_CA11N_373","v_CA11N_376","v_CA11N_379","v_CA11N_382")
vectors=c("v_CA16_3636","v_CA16_3639","v_CA16_3669","v_CA16_3699","v_CA16_3741","v_CA16_3810","v_CA16_3750","v_CA16_3756","v_CA16_3783")

We choose the categories and colours we want to map and define a convenience function to rename the variables and compute the qantities for the other asian countries that we don’t break out.

categories=c("Americas","Europe","Africa + Oceania","Philippines","China","India","Other Asian Countries")
colors=c("#7a0177", "#3333cc", "#ff00ff", "#00ffff", "#ff1a1c", "#4dff4a", "#ffff33")
prep_data <- function(geo){
  data <- geo@data %>% replace(is.na(.), 0)
  data <- rename(data,
    total=v_CA16_3636,
    Americas=v_CA16_3639,                                  
    Europe=v_CA16_3669,
    Philippines=v_CA16_3783,
    China=v_CA16_3750,
    India=v_CA16_3756)
  data <- mutate(data,`Other Asian Countries` = v_CA16_3741-Philippines-China-India,
                     `Africa + Oceania`=v_CA16_3699 + v_CA16_3810)
  geo@data <- data
  return(geo)
}
prep_data_2011 <- function(geo){
  data <- geo@data %>% replace(is.na(.), 0)
  data <- rename(data,
    total=v_CA11N_265,
    Americas=v_CA11N_268,                                  
    Europe=v_CA11N_304,
    Africa=v_CA11N_334,
    Philippines=v_CA11N_376,
    China=v_CA11N_379,
    India=v_CA11N_382)
  data <- mutate(data,`Other Asian Countries` = v_CA11N_373-Philippines-China-India)
  geo@data <- data
  return(geo)
}

Next we grab the data via cancensus,

data_csd=get_census(dataset = 'CA16', regions=regions,vectors=vectors,geo_format='sp',labels='short',level='CSD') %>% prep_data

OGR data source with driver: GeoJSON 
Source: "/Users/jens/.cancensus_cache/CM_geo_6b42dc9bcf266cb7c9bfbbbf40889c49.geojson", layer: "OGRGeoJSON"
with 39 features
It has 11 fields

data_ct=get_census(dataset = 'CA16', regions=regions,vectors=vectors,geo_format='sp',labels='short',level='CT') %>% prep_data

OGR data source with driver: GeoJSON 
Source: "/Users/jens/.cancensus_cache/CM_geo_4a5626a6774fefc22c656038b44803a8.geojson", layer: "OGRGeoJSON"
with 478 features
It has 11 fields

data_da=get_census(dataset = 'CA16', regions=regions,vectors=vectors,geo_format='sp',labels='short',level='DA') %>% prep_data

OGR data source with driver: GeoJSON 
Source: "/Users/jens/.cancensus_cache/CM_geo_e0f185a0c335b032df17fa063ccc9238.geojson", layer: "OGRGeoJSON"
with 3450 features
It has 10 fields

data_db=get_census(dataset = 'CA16', regions=regions,geo_format='sp',labels='short',level='DB')

OGR data source with driver: GeoJSON 
Source: "/Users/jens/.cancensus_cache/CM_geo_ebd06436dd29d8db88d687516f5a8c04.geojson", layer: "OGRGeoJSON"
with 15197 features
It has 10 fields

which we then re-aggregate to make sure we don’t miss overall counts due to privacy cutoffs distribute them proportionally among the population.

data_ct@data <- dot_density.proportional_re_aggregate(data=data_ct@data,parent_data=data_csd@data,geo_match=setNames("GeoUID","CSD_UID"),categories=categories,base="Population")
data_da@data <- dot_density.proportional_re_aggregate(data=data_da@data,parent_data=data_ct@data,geo_match=setNames("GeoUID","CT_UID"),categories=categories,base="Population")
data_db@data <- dot_density.proportional_re_aggregate(data=data_db@data,parent_data=data_da@data,geo_match=setNames("GeoUID","DA_UID"),categories=categories,base="Population")

Map

All that’s left to do is to covert our re-aggregated block-level data to dots, using the dot_density.compute_dots function from the dotdensity package and feed it into the dot_density.dots_map function to add them to our basemap.

# 1 dot = 5 immigrants
scale=5
dots.db <- dot_density.compute_dots(geo_data = data_db, categories = categories, scale=scale)
basemap +
  # zoom in a bit
  coord_fixed(xlim=c(-123.29,-122.6), ylim=c(49.02,49.35), ratio = 1/cos(49.2/180*pi)) +
# shade unpopulated blocks
#  geom_polygon(data=data_db[data_db$Population<=5,], 
#                        aes(long, lat, group = group), 
#                        fill = "#222222", size=0.1, 
#                        color = "#222222") +
  scale_colour_manual(values = colors) +
  labs(color = "",
                title="Immigrants 2011 - 2016",
                caption="Source: StatCan Census 2016 via cancensus & CensusMapper.ca",
                subtitle = paste0("1 dot = ",scale," people")) + 
  dot_density.dots_map(dots=dots.db,alpha=0.75,size=0.25)

# save image for later
ggsave('../images/recent_immigrants.png',width=26,height=26)